ADALINE
Adaline
- Date
- 1960 · confidence high
- Place
- Stanford, California, United States
- Credited to
- Marcian Hoff y Bernard Widrow
- Field
- Knowledge & symbols
What it is
ADALINE is one of the earliest single-layer artificial neural networks, based on the perceptron, whose weights were physically implemented with rheostats; it was widely used in adaptive signal processing, such as adaptive noise filtering.
What it led to
- adaptive signal processing
- adaptive noise filtering in telecommunications
- development of the LMS (Widrow-Hoff) learning algorithm
- evolution toward multilayer neural networks